Coming Q1 2017

Knuscan is our passive speaker authentication portfolio, using natural speech to validate a user. Since Knuscan uses natural interaction without any specific prompting, authentication can be transparent to the end user. A voice model can be created with 30 seconds of speech that can be captured during a natural conversation, while providing voice commands to an IoT device, or in a formal enrollment. Authentication then takes as little as 5 seconds of speech and increases in accuracy with longer utterances. Knuscan products are ideal for dropping into existing voice applications to increase security, to personalize an interaction, or to acquire metadata about the interaction.

Uses cases for Knuscan are as prevalent as the customer and employee interactions. Enterprise use cases include: Streamlining call center authentication; confirming a user’s identity throughout a phone call or interaction; or scanning audio for specific people for Edisovery. IoT use cases include: Smart Home and Smart Car interactions that leverage voice commands, and interaction with Smart Phones, Smart Watches, etc. Knuscan can streamline those experiences so consumers receive faster, secure and more personalized experiences.

Background

Passive speaker verification allows users to verify their identity with their natural day to day speech. Background can work with any language or channel, and like all KnuVerse products is designed to work reliably in noise. Flexible in deployment, it can be used within the context of call centers, apps, or IoT devices.

Background needs only 30 seconds of initial speech to create a voiceprint of an individual. Once that voiceprint is created, Background needs only 5 seconds of speech to verify a users identity.

Enrollment can be performed in a number of ways depending on the use case: Speech segments from a normal conversation, commands given to a device, or a formal enrollment can be used based on administrator / customer preference. KnuVerse works with enterprises to help them craft the enrollment methodology that is best for their use case and requirements.

Background provides a confidence score which allows for tailoring of the service to the sensitivity of each application. Watch the video to learn more!

DNA

DNA uses the power of natural language to detect when a specified person is speaking. DNA is used for a broad range of applications, from enabling different levels of access to users in a Smart Car or a Smart Home IoT application, to security alerting on a blacklist within a communications system.

DNA uses KnuVerse’s military grade voice biometrics, and simply requires 30 seconds of audio of an individual to create a voiceprint. The 30 seconds of audio doesn’t even have to be continuous! After enrolling, DNA will flag with specified level of certainty whether an enrolled individual is present in the audio.

Enrollment can be performed in a number of ways depending on the use case. Speech segments from a normal conversation, commands given to a device, or a formal enrollment can all be used to create the model of an individual. KnuVerse works with enterprises to help them craft the enrollment methodology that is best for their use case and requirements.

Check back soon to see demos of DNA in action!